Interference arises when the treatment assigned to one individual affects the outcomes of other individuals. Commonly, individuals are naturally grouped into clusters, and interference occurs only among individuals within the same cluster, a setting referred to as partial interference. We study network causal effects on outcome quantiles in the presence of partial interference. We develop a general nonparametric efficiency theory for estimating these network quantile causal effects, which leads to a nonparametrically efficient estimator. The proposed estimator is consistent and asymptotically normal with parametric convergence rates, while allowing for flexible, data-adaptive estimation of complex nuisance functions. We leverage a three-way cross-fitting procedure that avoids direct estimation of the conditional outcome distribution. Simulations demonstrate adequate finite-sample performance of the proposed estimators, and we apply the methods to a clustered observational study.

The Bell colouring graph $\mathcal{B}(G)$ of a graph $G$ is the graph whose vertices are the partitions of the vertex set of $G$ into independent sets, with an edge between two partitions if and only if one can be obtained from the other by changing the part of a single vertex of $G$. Given a natural number $k$, the Bell $k$-colouring graph $\mathcal{B}_k(G)$ and the upper-Bell $k$-colouring graph $\mathcal{B}_{\geq k}(G)$ are the induced subgraphs of $\mathcal{B}(G)$ consisting of all partitions with at most $k$ parts and at least $k$ parts, respectively. We determine precisely when two finite graphs have isomorphic Bell colouring graphs. In particular, we show that every $n$-vertex graph $G$ with no vertices of degree $n-1$ is uniquely determined by its Bell colouring graph $\mathcal{B}(G)$, and by its upper-Bell colouring graph $\mathcal{B}_{\geq k}(G)$ if $k\leq n-2$. We also show that every $n$-vertex graph with maximum degree $Δ(G)< \frac{1}{9}n-\frac{1}{3}$ is uniquely determined by its Bell $k$-colouring graph $\mathcal{B}_k(G)$ if $k>χ(G)$. By taking graph complements, each of these results can be restated in terms of partitions into cliques.

We investigate the cosmological implications of non-polynomial quasi-topological gravity (NPQTG), a novel class of modified gravitational theories in which the background dynamics is encoded in a single function of the Hubble parameter. This framework provides a minimal and theoretically consistent extension of general relativity, incorporating higher-curvature effects while preserving second-order field equations and avoiding higher-derivative instabilities. We first establish the general conditions for cosmological viability and construct explicit realizations, including polynomial, quartic, power-law and non-polynomial models, demonstrating how different functional forms lead to distinct expansion histories. Focusing on the quartic and power-law cases, we show that the resulting cosmological evolution reproduces the standard thermal history of the Universe and gives rise to an effective dark-energy sector of geometric origin, with dynamical equation-of-state behavior that can lie in the quintessence or phantom regime. We then confront the models with observational data from Type Ia Supernovae, Cosmic Chronometers, and Baryon Acoustic Oscillations, using a Bayesian MCMC analysis. We find that both models provide an excellent fit to the data, remaining fully compatible with current constraints and statistically competitive with $Λ$CDM. Our results demonstrate that NPQTG offers a simple and efficient framework for describing late-time cosmic acceleration with dynamical dark energy, while maintaining theoretical consistency and observational viability.

Little Red Dots (LRDs) are a population of compact, red sources that have emerged as one of the most puzzling findings of JWST. Variability provides a direct probe of their central engines. Here we present the first joint spectroscopic and photometric time-domain study of LRDs undertaken with the JWST TWINKLE slitless spectroscopy program. Surveying the FRESCO GOODS-North legacy field, TWINKLE monitors a complete, H$α$-flux-limited sample of 18 LRDs at z = 3.9-6.8, achieving a rest-frame baseline of $\sim$140-220 days. We detect no variability in photometry, H$α$ line flux, or line shape across the sample. If LRDs resembled AGN in reverberation mapping samples -- the foundation for black hole mass calibrations and luminosity scaling relations -- we would expect >10 sources to show measurable fluctuations. Observing none implies a 5.9$σ$ deficit. The non-detections hold across all broad H$α$ emitters within TWINKLE's field of view -- the 18 V-shaped LRDs as well as 9 non-LRDs. Comparison with simulated light curves disfavors sub-Eddington accretion and is instead consistent with super-Eddington accretion, other mechanisms that suppress variability, or perhaps no AGN whatsoever. If LRDs do harbor black holes, calibrations derived from sub-Eddington systems may not apply, thereby explaining JWST's apparently "overmassive" black holes. These observations provide unique constraints on the physics of one of the most enigmatic populations discovered by JWST.

Large Language Models (LLMs) have shown much promise in powering a variety of software engineering (SE) tools. Offering natural language as an intuitive interaction mechanism, LLMs have recently been employed as conversational ``programming assistants'' capable of supporting several SE activities simultaneously. As with any SE tool, it is crucial that these assistants effectively meet developers' needs. Recent studies have shown addressing this challenge is complicated by the variety in developers' needs, and the ambiguous and unbounded nature of conversational interaction. This paper discusses our current and future work towards characterizing how diversity in cognition and organizational context impacts developers' needs, and exploring personalization as a means of improving the inclusivity of LLM-based conversational programming assistants.

We establish sufficient conditions on discrete subsets of $\mathbb{R}^d$ for them to form a uniqueness or a non-uniqueness pair for the fractional Laplacian. Specifically, assuming that $f=0$ on $Λ$ and that $(-Δ)^sf=0$ on $M$, where $Λ, M \subset \mathbb{R}^d$ are discrete, we find sufficient conditions on these sets that force $f$ to vanish identically, and we provide examples in which non-uniqueness occurs. Some of the ideas used in the proofs also extend to a broader class of multiplier operators.

This work establishes that an optimal transport~(OT) problem regularized by a given $f$-divergence admits the same solution as another OT problem regularized by a different $g$-divergence, under an appropriate transformation of the cost function. This structural equivalence between OT problems regularized by distinct divergences, in the sense of sharing the same unique minimizer, is demonstrated within the framework of Polish spaces with bounded cost functions.

Supersonic winds from massive stars carry great amounts of kinetic power and modify the surrounding interstellar medium. Through this interaction a stellar bubble is formed. Theoretical studies and recent observations suggest that the winds of massive stars could be sources of Galactic cosmic rays. The first detection of synchrotron emission from the bubble of a single star was reported, indicating the presence of relativistic electrons. Studying the non-thermal emission from a single massive star can help to better understand the acceleration of particles taking place in massive star clusters. WR 102 is the perfect case of study. In this work, we present the first high-energy model for the bubble of WR 102: G2.4+1.4. We aim at fitting the radio data and predicting gamma-ray emission. We assume that both electrons and protons are accelerated at the wind shock. We applied a classical model for the stellar bubble and adopted a one-zone model for estimating the radiation produced by the relativistic particles near the acceleration region. Additionally, we computed the expected emission from the protons that diffuse to the outer regions of the bubble. Also, we estimated the leptonic and hadronic contributions expected from cosmic rays. We fitted the observations considering that 3% of the wind kinetic power goes into relativistic electrons, and a magnetic field of 250 $μ$G. The dominant component at high energies is produced by locally accelerated protons reaching the shell. Protons might reach PeV energies in the wind bubble, but the predicted gamma-ray flux is too low to be detectable.

This study analyzes the impacts of exchange rate and inflation on exports in Turkiye. Annual data for the period 1995-2023 were used in the analysis. The Johansen cointegration analysis and Dynamic Least Squares (DOLS) method were employed in the study. Identifying the cointegration relationship enabled the estimation of the long-run coefficients. The results show that an increase in the real effective exchange rate (appreciation of the Turkish lira) and inflation reduce exports with coefficients of -0.185 and -0.125, respectively. Foreign direct investment and imports, added to the study as control variables, have a positive impact on exports with coefficients of 0.117 and 0.849, respectively. These findings indicate that exchange rate stability and inflation control are priorities for improving foreign trade performance. Furthermore, policies that increase foreign direct investment and strategically manage imports complement this process.

Item cold-start is a pervasive challenge for collaborative filtering (CF) recommender systems. Existing methods often train cold-start models by mapping auxiliary item content, such as images or text descriptions, into the embedding space of a CF model. However, such approaches can be limited by the fundamental information gap between CF signals and content features. In this work, we propose to avoid this limitation with purely content-based modeling of cold items, i.e. without alignment with CF user or item embeddings. We instead frame cold-start prediction in terms of item-item similarity, training a content encoder to project into a latent space where similarity correlates with user preferences. We define our training objective as a sparse generalization of sampled softmax loss with the $α$-entmax family of activation functions, which allows for sharper estimation of item relevance by zeroing gradients for uninformative negatives. We then describe how this Sampled Entmax for Cold-start (SEMCo) training regime can be extended via knowledge distillation, and show that it outperforms existing cold-start methods and standard sampled softmax in ranking accuracy. We also discuss the advantages of purely content-based modeling, particularly in terms of equity of item outcomes.

We investigate how strongly late-time inferences about DE dynamics depend on the functional prior used to represent the expansion history. Using identical late-time combinations of CC, DESI BAO measurements, the Pantheon+ SN1a sample, and the H0DN prior, we compare a node-based reconstruction of the reduced Hubble function $E(z)$ with a representative family of smooth low-dimensional DE EoS parametrizations, including CPL. Over the redshift range constrained by the data, both approaches yield consistent $H(z)$, and, in the absence of H0DN, compatible values of $H_0$. However, a clear method dependence emerges at intermediate redshift ($z\sim1.7$): the reconstruction favors stronger deceleration, $q_{\rm Rec}(1.7)\simeq0.56-0.61$, whereas the smooth parametrizations cluster at $q(1.7)\simeq0.32-0.40$, implying a persistent $\sim2-3σ$ discrepancy across dataset combinations and parametrizations. For the EoS-based parametrizations, whose effective DE densities remain positive by construction, the preferred $w_{\rm DE}(1.7)<-1$ values correspond to NECB-violating (phantom-like) behaviour, but this is a less robust discriminator as $w_{\rm DE}$ becomes ill-conditioned as $ρ_{\rm DE}\to0$. In the effective-fluid mapping, the reconstruction accommodates the same late-time kinematical preference through a rapid descent of $ρ_{\rm DE}(z)$ toward very small values and a sign change, whereas the EoS-based parametrizations absorb it through smoother, and in several cases NECB-violating, evolution over $z\sim1-2$. Although the reconstruction improves the best-fit likelihood, especially with H0DN, Bayesian evidence continues to favor the simpler parametric descriptions. Our results isolate $z\sim1.5-2$ as the key window in which EoS-based DE parametrizations can compress localized kinematic structure and associated features of DE that are still permitted by current late-time data.

The emergence of Cryptographically Relevant Quantum Computers (CRQCs) presents a critical threat to classical cryptographic systems, particularly widely adopted protocols such as RSA, Diffie-Hellman (DH), and Elliptic Curve Cryptography (ECC). Given their extensive use in the financial sector, the advent of quantum adversaries compels banking institutions to proactively develop and adopt quantum-safe communication mechanisms. This paper introduces a hybrid quantum-safe architecture, orchestrated via Software-Defined Networking (SDN) key distribution. The proposed framework enables the early integration of Classical Cryptography (CC), Quantum Key Distribution (QKD), and Post-Quantum Cryptography (PQC) within a Dynamic Multipoint Virtual Private Network (DMVPN) environment, providing highly scalable, full-mesh, site-to-site encrypted communications for enterprise networks. This is particularly relevant at a time when PQC algorithms have not yet been incorporated into finalized IPsec standards. The architecture has been validated across a five-node testbed comprising three physical nodes within a campus network in Madrid and two private-cloud nodes located in the north of Spain and Mexico. The deployment leverages a heterogeneous mix of physical and virtual devices, diverse technology providers, Discrete Variable QKD (DV-QKD) and Continuous Variable QKD (CV-QKD) implementations, and mutually incompatible key-delivery interfaces (ETSI004, ETSI014 and Cisco SKIP), demonstrating flexibility, scalability, and interoperability across environments. Through this framework, we demonstrate that quantum-safe communication in financial networks is not only technically feasible but also scalable, interoperable, and resilient. The proposed architecture establishes a robust, flexible, and future-proof foundation for secure financial communications in the era of quantum computing.

We develop a variationally consistent mesoscopic extension of Cosserat elasticity motivated by the breakdown of compatibility in classical formulations. By admitting compatibility-breaking perturbations, the classical theory ceases to remain closed under admissible variations, necessitating an enlargement of the constitutive framework. This leads naturally to a formulation in which torsion and curvature are treated as independent distributed measures of defects. The theory is constructed using a Palatini-type variational approach, with the coframe and connection as independent fields. The resulting Euler--Lagrange equations yield both the standard balance laws and defect-related excitation fields. Material invariance gives rise to configurational forces and moments, which emerge as Noether currents and are directly linked to defect transport governed by the Bianchi identities. The framework provides a unified description of defect kinematics, configurational mechanics, and microstructural evolution. Illustrative examples and numerical evaluations demonstrate how defect transport generates configurational forces and highlight the underlying Maxwell-type structure of the theory. The proposed formulation offers a consistent geometric foundation for the analysis of structured solids with evolving internal geometry and provides a basis for future developments in defect dynamics and dissipative processes.

Recently, the authors of this paper introduced logarithmic Hochschild (co)homology of logarithmic spaces in a geometric way using formality of derived intersections. In this paper, the authors extend the decomposition theorem for the logarithmic Hochschild (co)homology of firm orbifolds to general logarithmic orbifolds and consider two applications of the decomposition theorem. First, we consider two versions of a symmetric product and compute the logarithmic Hochschild homology of them. Second, we show that logarithmic Hochschild homology is invariant under root stack operations.

Martinez-Rivillas and de Queiroz gave extensional Kan semantics for the untyped lambda-calculus and later constructed the concrete K-infinity homotopy-model. The two main mathematical results of the present paper are these. First, we show that a smaller front-seed coherence package (WL, WR) together with an inner-right-front pentagon contraction already suffices to recover the associator comparison, semantic pentagon, and bridge theorems used in the later semantic arguments. Second, we prove explicit global reify, reflect, and application formulas for K-infinity, with exact coordinatewise identities at every finite stage. We also record two structural clarifications: the recursive all-dimensional continuation of the explicit low-dimensional tower is obtained by a finite packaging phase followed by a uniform equality-generated recursion; and, on a deliberately fixed forward witness language for the classical separation span, the canonical identity-type higher tower on K-infinity forces all higher non-connection once the two witness classes land at distinct points. The paper is fully formalized in Lean 4, and the project sources contain no local uses of sorry, admit, or axiom.

It is known that the semiclassical approximation to the gravity path integral can be leveraged to explain certain inherently quantum aspects of gravity. One such aspect is the state-counting interpretation of the Bekenstein-Hawking entropy of black holes. A second aspect is the Page curve for the entanglement entropy of Hawking radiation, which agrees with expectations from unitarity. We show that these two questions are equivalent from the gravity path integral point of view. In particular, the Hawking's information loss puzzle gets resolved automatically by considering any (over)complete basis of black hole microstates which is compatible with black hole entropy. The tool which relates the two questions is a convex optimization problem for the von Neumann entropy of Hawking radiation.

Comparison of the measured spectral emission properties of single-crystal Cu(001) and W(111) photocathodes to established photoemission theories reveal evidence for an additional one photon emission process predominantly affecting electron emission near and below the photoemission threshold. This additional photoemission process is postulated to be due to a momentum-resonant Franck-Condon mechanism mediated by inelastic Umklapp electron scattering. An initial first-principles simulation of this emission process (involving the electron thermal effective mass, the inelastic electron mean free path at the vacuum level, and the number of Fermi surfaces in the metal), when combined with a direct one-step band emission model, is consistent with the measured spectral dependencies of both the quantum efficiency and mean transverse energy of electron photoemission from the two single-crystal metal photocathodes.

We define dynamic treatment regimes and associated potential outcomes for data described by marked point processes (MPPs). These definitions motivate MPP analogues of the commonly used consistency, exchangeability, and positivity conditions that are sufficient for identifying effects in MPP data structures. The conditions are formulated based on martingale theory, which allows us to derive explicit identifying assumptions for data described by stochastic processes. The definitions and conditions align with well-established discrete-time results in important special cases. Thus, this work bridges the large literatures on survival (event history) analysis with counting processes in continuous time and causal inference with variables in discrete-time. After formulating a set of identification conditions, we derive and characterize marginal g-formulas. The g-formulas are generally different from those studied in related works, though they coincide in important special cases. We relate our findings to previous work on causal inference with (counting) processes, the classical survival literature, and the discrete-time causal inference literature.

Through semiclassical methods the subject of quantum chaos motivates and depends on Hamiltonian chaos research. Presented here is a selection of Hamiltonian chaos topics that in this way get directly related to any of a variety of quantum chaos research problems. The chapter begins with a description of various useful theoretical and computational tools of chaos research, e.g.~surfaces of section, paradigms of chaos, stability analysis, and symbolic dynamics... This is followed by discussions regarding the geometry of chaos, how chaotic systems respond to perturbations, and the complexification of Hamiltonian dynamics. The emphasis is on intuitive explanations and illustrations of various ideas with the references containing more mathematically rigorous expositions.

We study the dynamical spin and nematic correlations in the bilinear-biquadratic spin-1 chain in the critical phase hosting deconfined spinons. We demonstrate how spinon interactions can be directly probed in the presence of a magnetic field or a single-ion anisotropy. Our analytical predictions are supported by numerical matrix-product-state (MPS) simulations of the underlying microscopic model.

Open Radio Access Network (O-RAN) is an important 5G network architecture enabling flexible communication with adaptive strategies for different verticals. However, testing for O-RAN deployments involve massive volumes of time-series data (e.g., key performance indicators), creating critical challenges for scalable, unsupervised monitoring without labels or high computational overhead. To address this, we present ESN-DAGMM, a lightweight adaptation of the Deep Autoencoding Gaussian Mixture Model (DAGMM) framework for time series analysis. Our model utilizes an Echo State Network (ESN) to efficiently model temporal dependencies, proving effective in O-RAN networks where training samples are highly limited. Combined with DAGMM's integratation of dimensionality reduction and density estimation, we present a scalable framework for unsupervised monitoring of high volume network telemetry. When trained on only 10% of an O-RAN video-streaming dataset, ESN-DAGMM achieved on average 269.59% higher quality clustering than baselines under identical conditions, all while maintaining competitive reconstruction error. By extending DAGMM to capture temporal dynamics, ESN-DAGMM offers a practical solution for time-series analysis using very limited training samples, outperforming baselines and enabling operator's control over the clustering-reconstruction trade-off.

Inspired by the structure of $srs$ Laves networks in $\mathbb{R}^3$ that underpin the celebrated gyroid surface, we construct a Laves network of identical three-coordinated vertices on $S^3$ with double-twist. This network is a subset of the vertices and edges of the 600-cell, and can be viewed as a bipartite graph of disjoint 24-cell vertices inscribed in the 600-cell. We describe mutually entangled realizations of this network on $S^3$, and describe their relation to the well-known $srs$ Laves network structure in $\mathbb{R}^3$.

The increase in data volume, computational resources, and model parameters during training has led to the development of numerous large-scale industrial retrieval models for recommendation tasks. However, effectively and efficiently deploying these large-scale foundational retrieval models remains a critical challenge that has not been fully addressed. Common quick-win solutions for deploying these massive models include relying on offline computations (such as cached user dictionaries) or distilling large models into smaller ones. Yet, both approaches fall short of fully leveraging the representational and inference capabilities of foundational models. In this paper, we explore whether it is possible to learn a hierarchical organization over the memory of foundational retrieval models. Such a hierarchical structure would enable more efficient search by reducing retrieval costs while preserving exactness. To achieve this, we propose jointly learning a hierarchical index using cross-attention and residual quantization for large-scale retrieval models. We also present its real-world deployment at Meta, supporting daily advertisement recommendations for billions of Facebook and Instagram users. Interestingly, we discovered that the intermediate nodes in the learned index correspond to a small set of high-quality data. Fine-tuning the model on this set further improves inference performance, and concretize the concept of "test-time training" within the recommendation system domain. We demonstrate these findings using both internal and public datasets with strong baseline comparisons and hope they contribute to the community's efforts in developing the next generation of foundational retrieval models.

A screening of oxygen profiles in mid-T treated SRF cavities is crucial, in order to infer physical correlations between the microscopic cavity lattice and cavity performance - a problem concerning acceleration physicists for years. This thesis provides an analysis of oxygen diffusion profiles for three differently treated samples: Two mid-T baked and the third with the standard EuXFEL recipe. The measurement method utilizes EXAFS spectroscopy and was carried out at the DELTA facility in Dortmund. The result suffers heavily under noise, making the quantity of the result barely useable. Qualitatively, no deviations of current models regarding the profiles, could be proven, and no results of previous studies were contradicted. The experimental analysis is described in precision, interpretations of the possible are undertaken and theoretical considerations regarding error estimation and possible EXAFS simulations for future attempts are provided. A repetition at the PETRA III facility is indicated.

For stochastic Hamilton-Jacobi (SHJ) equations, instability points are the space-time locations where two eternal solutions with the same asymptotic velocity differ. Another fundamental structure in such equations is shocks, which are the space-time locations where the velocity field is discontinuous. In this work, we study the KPZ fixed point, the central object of the KPZ universality class, which can be viewed as a prototype--albeit degenerate--of an inviscid SHJ equation in one spatial dimension. We describe the geometric structure of the instability region and give a detailed and precise analysis of its interplay with the shock structures of the two eternal solutions. We show that these shock structures allow one to reconstruct the instability region. Along the way, we obtain a complete classification of all possible configurations of semi-infinite geodesics emanating from arbitrary space-time points, in the directed landscape--the random environment in which the KPZ fixed point evolves.

Achieving consistent time across devices in distributed systems often involves exchanging timestamped messages over a network. Precise time synchronization is crucial for applications such as cellular networks, industrial automation, and transactional databases. However, delay variation in synchronization packets-often caused by congestion from competing traffic-degrades synchronization accuracy. Detecting whether a packet experienced congestion can help improve synchronization through filtering and statistical methods. We propose an in-network congestion indication and filtering mechanism for synchronization messages used in protocols such as the Network Time Protocol (NTP) and Precision Time Protocol (PTP). Network devices mark packets that experienced queuing, allowing clocks to correct errors caused by varying delays. Our approach requires only simple changes at switches or routers, avoiding deep packet inspection or protocol modifications. The method is backward compatible, using standard but currently unused fields in IP, PTP, or NTP headers. We implement our method on a Tofino P4 target and demonstrate an improvement of over 80% in synchronization performance over a single hop. Moreover, we show that the performance of traditional statistical filters, such as min-RTT and median-delay, is improved by 90% over the one-hop hardware setup. We further demonstrate the effectiveness of our proposed method across multiple hops, both analytically and through simulation. Congestion marking improves the root-mean-squared clock offset estimation error by 30% to 80%, depending on network conditions and filtering techniques.

This paper investigates the robust stability problem of a feedback system in the presence of uncertainties induced by graphical regions in the plane where the scaled relative graphs (SRGs) reside. Our main results are developed using a novel and intuitive concept, the Davis-Wielandt shell, together with its connection to SRGs and related variants. We first study a matrix robust nonsingularity (MRN) problem for two types of graphically induced uncertainty sets: one with prior information on $θ$ and one without. In the former case, we show that, whenever the uncertainty-inducing region is mirror symmetric about the $θ$-axis, the separation between a specific variant of the SRG and the region provides a necessary and sufficient condition for MRN. When the region is asymmetric, the necessity generally fails. This recovers the necessity of the small gain condition, and reveals the necessity of small angle conditions and sectored-disc conditions at the matrix level. In the latter case, we show that an additional $θ$-circular connectivity property is required to obtain necessary and sufficient conditions. Building on these MRN results, we then derive sufficient conditions for robust stability of multi-input multi-output (MIMO) linear time-invariant (LTI) systems under frequencywise symmetric uncertainties. In addition, connections with existing system characteristics such as disc-boundedness are discussed and exploited to obtain state-space characterisations for angle-bounded and mixed gain-angle-bounded systems. Based on these results, we construct a $θ$-angle-gain profile of a system that provides an intuitive visualisation of its feedback robustness against conic and sectorial uncertainties.

Open Radio Access Network (O-RAN) architectures enhance flexibility for 6G and NextG networks. However, it also brings significant challenges in O-RAN testing with evaluating abundant, high-dimensional key performance indicators (KPIs). In this paper, we introduce a novel two-stage framework to learn temporally-aware low-dimensional representations of O-RAN testing KPIs. To be specific, stage one employs an information-theoretic H-score to train a hybrid self-attentive transformer and echo state network (ESN) reservoir, called Transformer-ESN, capturing temporal dynamics and producing task-aligned $8$-dimensional embeddings. Stage two evaluates these embeddings by training a lightweight multilayer perceptron (MLP) predictor exclusively on them for key target KPIs such as reference signal received quality (RSRQ) and spectral efficiency. Using real-world O-RAN testbed data (video streaming with interference), our approach demonstrates a significant advantage specifically when training samples are very limited. In this scenario, the low-dimensional representations learned from the Transformer-ESN yield mean square error (MSE) reductions of up to 41.9\% for RSRQ and 29.9\% for spectral efficiency compared to predictions from the original high-dimensional data. The framework exhibits high efficiency for O-RAN testing, significantly reducing testing complexities for O-RAN systems.

This paper examines the polarization-dependent output of various types of 3-mode photonic lanterns fabricated using double-clad fibers. We explore the sensitivity of the modal response across several types of photonic lanterns, from the fully symmetric and strongly coupled structure of regular photonic lanterns to the fully asymmetric structure of mode-selective photonic lanterns. We demonstrate the high sensitivity of the output of photonic lanterns with strong coupling between their ports to the polarization of the input state. In contrast, ports with high isolation or low coupling, such as in mode-selective photonic lanterns, exhibit responses that are almost polarization independent.

In this paper, we investigate safety-critical control problem of discrete-time stochastic systems with incomplete information, where safety constraints must be enforced using state estimates obtained from noisy measurements. We develop an output-feedback control barrier function (CBF) framework based on an expectation-based discrete-time barrier condition that explicitly incorporates estimation uncertainty through the evolving belief over the state. To enable real-time implementation, we derive deterministic sufficient conditions that conservatively enforce the expectation-based CBF by bounding the expectation with computable functions of the belief statistics using Jensen inequalities. The resulting safety filter is formulated as a tractable optimization problem compatible with standard online controllers. Numerical simulations demonstrate that the proposed output-feedback approach achieves fast online computation while providing reliable safety performance in the presence of process noise and measurement uncertainty.